Laser-produced plasmas (LPPs) are formed in one example by irradiating Sn droplets with a focused laser beam. Because LPPs radiate in the extreme ultraviolet (EUV) range of the electromagnetic spectrum, they are considered to be a promising EUV radiation source for EUV lithography systems.
FIG. 1 is a schematic diagram of a generalized configuration for a prior art LPP-based source-collector module (“SOCOMO”) 10 that uses a normal-incidence collector (“NIC”) mirror MN, while FIG. 2 is a more specific example configuration of the “LPP-NIC” SOCOMO 10 of FIG. 1. The LPP-NIC SOCOMO 10 includes a high-power laser 12 that generates a high-power, high-repetition-rate laser beam 13 having a focus F13. LPP-NIC SOCOMO 10 also includes along an axis A1 a fold mirror FM and a large (e.g., ˜600 mm diameter) ellipsoidal NIC mirror MN that includes a surface 16 with a multi-layer coating 18. The multilayer coating 18 is essential to guarantee good reflectivity at EUV wavelengths. LPP-NIC SOCOMO 10 also includes a Sn source 20 that emits a stream of tin (Sn) pellets 22 that pass through laser beam focus F13.
In the operation of LPP-NIC SOCOMO 10, laser beam 13 irradiates Sn pellets 22 as the Sn pellets 22 pass through the laser beam focus F13, thereby produce a high-power LPP 24. LPP 24 typically resides on the order of hundreds of millimeters from NIC mirror MN and emits EUV radiation 30 as well as energetic Sn ions, particles, neutral atoms, and infrared (IR) radiation. The portion of the EUV radiation 30 directed toward NIC mirror MN is collected by the NIC mirror MN and is directed (focused) to an intermediate focus IF to form an intermediate focal spot FS. The intermediate focus IF is arranged at or proximate to an aperture stop AS. Only that portion of the EUV radiation 30 that makes it through aperture stop AS forms focal spot FS. Here is it noted that focus spot FS is not an infinitely small spot located exactly at intermediate focus IF, but rather is a distribution of EUV radiation 30 generally centered at the intermediate focus IF.
Advantages of LPP-NIC SOCOMO 10 are that the optical design is simple (i.e., it uses a single ellipsoidal NIC mirror) and the nominal collection efficiency can be high because NIC mirror MN can be designed to collect a large angular fraction of the EUV radiation 30 emitted from LPP 24. It is noteworthy that the use of the single-bounce reflective NIC mirror MN placed on the opposite side of LPP 24 from the intermediate focus IF, while geometrically convenient, requires that the Sn source 20 not significantly obstruct EUV radiation 30 being delivered from the NIC mirror MN to the intermediate focus IF. Thus, there is generally no obscuration in the LPP-NIC SOCOMO 10 except perhaps for the hardware needed to generate the stream of Sn pellet 22.
LPP-NIC SOCOMO 10 works well in laboratory and experimental arrangements where the lifetime and replacement cost of LPP-NIC SOCOMO 10 are not major considerations. However, a commercially viable EUV lithography system requires a SOCOMO that has a long lifetime. Unfortunately, the proximity of the surface 16 of NIC mirror MN and the multilayer coatings 18 thereon to LPP 24, combined with the substantially normally incident nature of the radiation collection process, makes it highly unlikely that the multilayer coating 18 will remain undamaged for any reasonable length of time under typical EUV-based semiconductor manufacturing conditions.
A further drawback of the LPP-NIC SOCOMO 10 is that it cannot be used in conjunction with a debris mitigation tool based on a plurality of radial lamellas through which a gas is flowed to effectively stop ions and neutrals atoms emitted from the LPP 24 from reaching NIC mirror MN. This is because the radial lamellas would also stop the EUV radiation 30 from being reflected from NIC mirror MN.
Multilayer coating 18 is also likely to have its performance significantly reduced by the build-up of Sn, which changes the critical reflective properties of the multilayer coating 18. Also, the aforementioned energetic ions, atoms and particles produced by LPP 24 will bombard multilayer coating 18 and destroy the layered order of the top layers of the multilayer coating 18. In addition, the energetic ions, atoms and particles will erode multilayer coating 18, and the attendant thermal heating from the generated IR radiation can act to mix or interdiffuse the separate layers of the multilayer coating 18.
While a variety of fixes have been proposed to mitigate the above-identified problems with LPP-NIC SOCOMO 10, they all add substantial cost and complexity to module, to the point where it becomes increasingly unrealistic to include it in a commercially viable EUV lithography system. Moreover, the Sn droplet LPP EUV light source is a complex and expensive part of the LPP-NIC SOCOMO 10. What is needed therefore is a less expensive, less complex, more robust and generally more commercially viable SOCOMO for use in an EUV lithography system that uses a simpler and more cost-effective LPP-based EUV radiation source.